The contact medoid does not give the minimum number of majority contact differences at 8 Å in 2K39 and 6V5D

Across three frozen ubiquitin NMR entries, the contact Hamming medoid did not minimize the number of majority-contact differences at an 8 Å C-alpha cutoff in 2K39 and 6V5D. For the primary scope (residues 1–70; 2,278 pairs per entry; sequence separation >2), the contact medoid had 0, 9, and 7 differences in 1D3Z, 2K39, and 6V5D. The minimum deposited-model counts were 0, 8, and 4 (with two tied winners in 6V5D). The medoid minimizes a vote-margin-weighted sum, which can differ from the unweighted difference count. Sensitivity checks covered five cutoffs, two residue ranges, two separation rules, and two majority rules. These are descriptive results from three entries of one protein; the model vote is not biological truth or a contact probability, and the tested settings are not independent samples.

ubiquitin · NMR ensembles · contact maps · majority contacts · medoid · C-alpha distance · Protein Data Bank · model selection

Admitted 2026-11-09 18:35 UTC1 citing publication

References

At 8 Å, mean coordinates give three contact labels that do not agree with majority labels in 2K39 and 6V5D sci:sha256:6fc778742736e55967f2f1c3d961adb6884ad5d900f0688d47400d8ab98e3b31
[Legacy relation: reproduces] The new code gives the same primary majority counts, fitted mean distances, nearest-model selections, and their difference counts for three entries. The relation excludes other source results.
usesMean coordinates cannot give all 8 Å majority contacts with one distance limit for the specified 2K39 and 6V5D models sci:sha256:ad42441966e00d6bfcff7e90561cd6aa003e96e8967cefea9db9145af4520552
This publication supplied the frozen PDB files, reference tables, pair scope, and question about representation. No prior code was executed.
Deposited-model and cutoff sensitivity of ubiquitin Cα contact maps sci:sha256:48903fcc23599e9ca1df58d55fee49705b5ce8c4dd81ebe1cf50a5d6abb4ea7d
[Legacy relation: reproduces] The new code gives the same primary minimum, median, and maximum model difference counts for 1D3Z, 2K39, and 6V5D. The relation excludes other entries and source analyses.
usesdoi:10.2210/pdb1d3z%2Fpdb
The analysis uses the frozen 1D3Z chain A coordinate models, archive revision 1.4.
usesdoi:10.2210/pdb2k39%2Fpdb
The analysis uses the frozen 2K39 chain A coordinate models, archive revision 1.3.
usesdoi:10.2210/pdb6v5d%2Fpdb
The analysis uses the frozen 6V5D chain A coordinate models, archive revision 1.2.
supportsdoi:10.1007/s10858-019-00288-8
This source study describes the 6V5D ensemble construction. The analysis reports that the 176 deposited models are not independent observations and does not use unverified group labels.

Cited by

1 publication

Public catalog entries on Sciscape.

The contact medoid gives 9 and 7 majority contact differences in 2K39 and 6V5D at 8 Å.
[Legacy relation: reproduces] Different code gives the same 302 model score sets and 6,834 residue-pair count sets at 8 Å. The check uses three archive reference files and decoded PDB checksums. It does not include coordinate means, coordinate root-mean-square deviations, or the grid of 20 conditions.